Satellite and radar fused isolated mature convective cloud cluster dynamic identification system and method

Abstract

The invention belongs to the technical field of artificial intelligence, and particularly provides an isolated mature convective cloud cluster dynamic identification system and method fusing satellites and radars, and the system comprises a data processing module which is used for carrying out the preprocessing of obtained wind cloud static meteorological satellite data and radar data; the convective cloud cluster sample extraction module is used for identifying convective cloud cluster samples; the isolated convective cloud cluster sample identification module is used for determining an isolated convective cloud cluster sample; the isolated mature convective cloud cluster sample screening module is used for carrying out mature convective and immature convective discrimination on the isolated convective cloud cluster samples and screening out the isolated mature convective cloud cluster samples; the sample dynamic inspection module is used for obtaining a final isolated mature convective cloud cluster sample; the identification model construction module is used for fusing and verifying satellite features and radar features of the final isolated mature convective cloud cluster sample and establishing an intelligent identification model; according to the technical scheme, full-life-cycle tracking of the isolated mature convection system can be realized.

Description

Technical Field

[0001] The present invention belongs to the field of artificial intelligence technology, and in particular relates to a system and method for dynamically identifying isolated mature convective clouds by integrating satellites and radars. Background Art

[0002] Severe convective weather, primarily characterized by short bursts of heavy rainfall, thunderstorms, strong winds, tornadoes, and hail, is one of the most frequent and devastating weather systems in summer. Accurate forecasts of severe convective weather are crucial for public safety, disaster reduction and mitigation, and military operations. Convective initiation (CI) signals the onset of severe convective weather activity and the initial phase of a severe convective storm, while isolated mature convection (IMC) signifies the mature stage of a convective system.

[0003] Severe convective weather is characterized by strong suddenness, small spatial scale, and short life cycle, especially the identification of isolated mature convection is more challenging. At present, the identification of local convection initiation (CI) mainly relies on satellite remote sensing and radar observations, but both have their limitations: satellite observations have the advantages of large-scale and long-term continuous monitoring, and can identify early convective signals (such as cloud optical thickness, brightness temperature changes, etc.) about 2 hours before convection is triggered, but it is difficult to accurately characterize the microphysical and dynamic characteristics of the cloud interior. Radar observations can provide high-precision cloud internal structure and precipitation information, such as the first appearance of the 35 dBZ echo can be used as a sign of heavy precipitation potential, but radar data is greatly affected by terrain obstruction, and for isolated and strong convective cells, its spatial resolution and scanning period may lead to monitoring lag. The problems of existing technologies mainly include two aspects: (1) Satellite and radar data are not fully integrated, and it is difficult to comprehensively utilize the macroscopic monitoring capability of satellites and the fine characterization capability of radars to accurately identify isolated mature convection. (2) There is uncertainty in the identification and tracking of convective cells, and it is impossible to accurately judge whether the convection will remain isolated or further organize and develop, which affects the accuracy of disastrous weather such as short-term heavy rainfall and thunderstorms.

[0004] Existing research uses fixed thresholds based on radar or satellite data to identify convective cloud clusters. However, fixed thresholds are difficult to adapt to the complexity of convection development under different weather conditions. Because different weather conditions, topography, humidity, and other factors affect the convective characteristics under the same threshold, some convection may not be identified (missed reports), while weak convection may be misclassified as strong convection (false alarms). In particular, the lack of dynamic tracking of the convective system life cycle makes it difficult to provide continuous short-term forecasts, resulting in short forecast timeliness. This approach ignores the nonlinear evolution of convective systems, namely that convection intensity may vary due to environmental conditions (such as wind shear and water vapor transport), rather than simple spatial displacement. Summary of the Invention

[0005] In order to solve the above problems, the application provides a satellite and radar fused isolated mature convective cloud cluster dynamic identification system and method, which combines new generation Fengyun geostationary meteorological satellite and dual-polarization weather radar data to monitor the whole life cycle from convective initial stage (CI) to mature stage, so as to solve the problem of low accuracy of isolated convective system.

[0006] The satellite and radar fused isolated mature convective cloud cluster dynamic identification system comprises:

[0007] A data processing module is configured to pre-process the acquired Fengyun geostationary meteorological satellite data and radar data.

[0008] A convective cloud cluster sample extraction module is configured to obtain a cloud mask of a convective cloud by using the pre-processed Fengyun geostationary meteorological satellite data, and identify a convective cloud cluster sample by using the cloud mask.

[0009] An isolated convective cloud cluster sample identification module is configured to determine an isolated convective cloud cluster sample based on the convective cloud cluster sample in combination with a cloud cluster growth region.

[0010] An isolated mature convective cloud cluster sample screening module is configured to use eight-connected region analysis to distinguish mature convective cloud from non-mature convective cloud for the isolated convective cloud cluster sample, and screen out an isolated mature convective cloud cluster sample.

[0011] A sample dynamic verification module is configured to verify a convective region of the isolated mature convective cloud cluster sample by using the pre-processed radar data, and obtain a final isolated mature convective cloud cluster sample.

[0012] An identification model construction module is configured to fuse satellite features and radar features of the verified final isolated mature convective cloud cluster sample, and construct an intelligent identification model.

[0013] An identification module is configured to complete dynamic identification of an isolated mature convective cloud cluster based on the intelligent identification model.

[0014] Preferably, the isolated convective cloud cluster sample identification module comprises:

[0015] A cold cloud nucleus identification unit is configured to identify a cold cloud nucleus by using an infrared channel in the pre-processed Fengyun geostationary meteorological satellite data.

[0016] A cloud cluster growth region acquisition unit is configured to check adjacent pixels of the cold cloud nucleus by using a growth search method, judge whether the adjacent pixels meet cloud cluster growth conditions, and obtain a cloud cluster growth region.

[0017] An isolated convective cloud cluster sample acquisition unit is configured to obtain an isolated convective cloud cluster sample based on the cloud cluster growth region.

[0018] Preferably, the sample dynamic verification module includes:

[0019] The radar echo corresponding cloud mask pixel acquisition unit is used to use the isolated mature convective cloud sample cloud mask extracted from the Fengyun geostationary meteorological satellite data, combine the binary function of the isolated mature convective cloud sample cloud mask and the radar echo area corresponding to the isolated mature convective cloud sample cloud mask, and obtain the number of pixels of the isolated mature convective cloud sample cloud mask corresponding to the radar echo area;

[0020] a cloud mask total pixel acquisition unit, configured to obtain the total number of pixels of the isolated mature convective cloud sample cloud mask by using a binary function of the isolated mature convective cloud sample cloud mask;

[0021] a matching rate calculation unit, configured to calculate matching rates between radar echo regions of different intensities and the isolated mature convective cloud sample mask using the number of pixels of the isolated mature convective cloud sample mask corresponding to the radar echo region and the total number of pixels of the isolated mature convective cloud sample mask;

[0022] The dynamic verification unit is configured to take the isolated mature convective cloud cluster sample whose matching rate meets a preset threshold as the final isolated mature convective cloud cluster sample.

[0023] Preferably, the recognition model building module includes:

[0024] A sample construction unit is used to verify the identification of the incipient convection of the final isolated mature convective cloud cluster samples using radar data, marking the final isolated mature convective cloud cluster samples that pass the verification as positive samples, and marking the isolated convective cloud cluster samples that fail to develop into mature convection as negative samples;

[0025] A feature extraction unit is used to extract feature parameters of Fengyun geostationary meteorological satellite data and radar data;

[0026] The model building unit is used to use the characteristic parameters to train the gradient boosting decision tree and build an intelligent recognition model.

[0027] Preferably, the process of using radar data to verify the inception of convection on the final isolated mature convective cloud cluster sample includes:

[0028] Time-matching of Fengyun geostationary meteorological satellite data with radar data using satellite-radar time matching spans of 15 minutes, 30 minutes, and 60 minutes.

[0029] When the satellite-radar time matching span is within 15 minutes, the convection is in the development stage, and the observations and the development stage threshold are used to infer the initial convection identification results of the previous 15 minutes; wherein the observations include reflectivity factor, differential reflectivity, correlation coefficient, and differential phase shift ratio;

[0030] When the satellite-radar time matching span is within 15-30 minutes, the convection is in the mature stage, and the observations and the mature stage threshold are used to infer the initial convection identification results of the first 30 minutes;

[0031] When the satellite-radar time matching span is within 60 minutes, the radar echo characteristics are used to verify the convection initiation identification results.

[0032] The present invention also provides a method for dynamically identifying isolated mature convective clouds by integrating satellite and radar, and the method includes:

[0033] Pre-process the acquired Fengyun geostationary meteorological satellite data and radar data;

[0034] Using pre-processed Fengyun geostationary meteorological satellite data, a cloud mask of convective clouds is obtained, and the convective cloud cluster samples are identified using the cloud mask;

[0035] Based on the convective cloud samples, and in combination with the cloud growth area, an isolated convective cloud sample is determined;

[0036] Eight-connected region analysis is used to distinguish mature convection from immature convection on the isolated convective cloud cluster samples, and isolated mature convective cloud cluster samples are screened out;

[0037] Using the pre-processed radar data to verify the convective area of ​​the isolated mature convective cloud cluster sample, and obtaining a final isolated mature convective cloud cluster sample;

[0038] The satellite and radar characteristics of the final isolated mature convective cloud cluster samples were integrated and verified to establish an intelligent recognition model;

[0039] Based on the intelligent recognition model, dynamic recognition of isolated mature convective cloud clusters is completed.

[0040] Preferably, the method for identifying the isolated convective cloud sample includes:

[0041] Cold cloud cores are identified using the infrared channel of pre-processed Fengyun geostationary meteorological satellite data.

[0042] Adopting a growth search method to check the adjacent pixels of the pixel corresponding to the cold cloud core, determining whether the adjacent pixels meet the cloud growth conditions, and obtaining the cloud growth area;

[0043] Based on the cloud growth area, isolated convective cloud samples are obtained.

[0044] Preferably, the method for obtaining the final isolated mature convective cloud sample comprises:

[0045] The isolated mature convective cloud sample mask extracted from Fengyun geostationary meteorological satellite data is used. The number of pixels of the isolated mature convective cloud sample mask corresponding to the radar echo area is obtained by combining the binary function of the isolated mature convective cloud sample mask and the radar echo area corresponding to the isolated mature convective cloud sample mask.

[0046] Using a binary function of the isolated mature convective cloud sample cloud mask, a total number of pixels of the isolated mature convective cloud sample cloud mask is obtained;

[0047] The matching rates between radar echo areas of different intensities and the isolated mature convective cloud sample masks are calculated using the number of pixels of the isolated mature convective cloud sample mask corresponding to the radar echo area and the total number of pixels of the isolated mature convective cloud sample mask.

[0048] The isolated mature convective cloud cluster sample whose matching rate meets a preset threshold is used as the final isolated mature convective cloud cluster sample.

[0049] Compared with the prior art, the present invention has the following beneficial effects:

[0050] The present invention improves the recognition and forecasting accuracy of isolated mature convective systems through three aspects: data fusion, dynamic tracking and intelligent identification: (1) Combining the large-scale cloud top information provided by geostationary meteorological satellites (FY4A, etc.) and the high temporal and spatial resolution reflectivity of ground-based Doppler radar, the joint identification of isolated mature convective systems is achieved. (2) Adaptive threshold and multi-parameter fusion methods are used to solve the misjudgment problem caused by a single data source and improve the recognition accuracy. (3) Through machine learning models, the evolution of radar echo intensity and satellite data characteristics are combined to achieve full life cycle tracking of isolated mature convective systems. Compared with traditional extrapolation methods, this method takes into account the nonlinear changes of convective systems and can accurately predict the strengthening, weakening or merging of convection, thereby improving the short-term forecasting capability. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the technical solution of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0052] Figure 1 This is a schematic diagram of the structure of a dynamic identification system for isolated mature convective clouds that integrates satellites and radars according to an embodiment of the present invention;

[0053] Figure 2 This is a detailed algorithm flow chart of the isolated mature convective cloud cluster dynamic identification system integrating satellite and radar according to an embodiment of the present invention. DETAILED DESCRIPTION

[0054] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0055] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0056] Example 1:

[0057] like Figure 1 、 Figure 2 As shown, the isolated mature convective cloud cluster dynamic identification system integrating satellite and radar includes: a data processing module, a convective cloud cluster sample extraction module, an isolated convective cloud cluster sample identification module, an isolated mature convective cloud cluster sample screening module, a sample dynamic verification module, an identification model construction module and an identification module.

[0058] The data processing module is used to preprocess the acquired Fengyun geostationary meteorological satellite data and radar data. In this embodiment, the satellite data mainly uses FY4A / AGRI L1 data, which is processed through longitude and latitude conversion, interpolation gridding, and calibration. The radar data is denoised and radial velocity correction is performed. The satellite data is interpolated at a 0.04° grid to match the horizontal resolution of the radar data. The nearest neighbor interpolation method is used to align the radar and satellite data in space. The Euclidean distance between the satellite and radar grid points is calculated, and the values ​​of the nearest neighbor points are matched, see formula (1), and the radar data at a time close to the satellite is selected.

[0059] , (1)

[0060] In the above formula, x and y represent the longitude and latitude of the satellite grid respectively. n and y n represent the longitude and latitude of the radar grid points respectively.

[0061] The convective cloud cluster sample extraction module is used to obtain the cloud mask of the convective cloud using the pre-processed Fengyun geostationary meteorological satellite data, and identify the convective cloud cluster samples using the cloud mask.

[0062] Specifically, based on the water vapor and infrared channels (6.25μm, 7.1μm, 10.7μm, and 12.0μm) from geostationary meteorological satellites, a cloud mask for convective clouds was designed. The convective cloud mask thresholds used are shown in Table 1:

[0063] Table 1

[0064]

[0065] Specifically, such as Figure 2 As shown in the figure, the convective system (CS) is tracked and identified based on the Fengyun-4 geostationary meteorological satellite. Convective clouds are identified using ground-based automatic weather station / precipitation data, based on the following criteria: BT10.8 < 0°C, BT7.1-BT10.7 > -15°C, BT12.0-BT10.7 > -5°C, and BT6.25-BT7.1 > -10°C.

[0066] The isolated convective cloud sample identification module is used to determine the isolated convective cloud samples based on the convective cloud samples and in combination with the cloud growth area.

[0067] Specifically, such as Figure 2 As shown in the figure, the brightness temperature (BT10.8) of the 10.8 μm channel of the FY4A / AGRI satellite is used to identify isolated convective cloud clusters. The judgment criteria are whether the BT10.8 temperature is less than or equal to 0°C and whether the cloud area is larger than 50 pixels (grid points). If so, the corresponding grid point pixels are removed. If not, the cloud top brightness temperature BTcloud(T) is obtained, and then the isolated convective cloud areas are extracted. These areas are then marked with connected domains to provide a basis for subsequent tracking and classification.

[0068] The isolated convective cloud sample identification module includes:

[0069] The cold cloud core identification unit is used to identify cold cloud cores using the infrared channel in the pre-processed Fengyun geostationary meteorological satellite data.

[0070] The cloud growth region acquisition unit is used to use the growth search method to check the adjacent pixels of the pixel corresponding to the cold cloud core to determine whether the adjacent pixels meet the cloud growth conditions and obtain the cloud growth region. Specifically, the lowest brightness temperature (BT) is first selected as the center of the cold cloud core, and its corresponding coordinates are (x0, y0). The adjacent pixels (x, y) are checked to determine whether they meet the cloud growth conditions:

[0071] , (2)

[0072] Where BT(x, y) is the brightness temperature of the current pixel, and BT(x0, y0) is the brightness temperature of the cold cloud core.

[0073] , (3)

[0074] In the above formula, k=2, is the standard deviation of the local brightness temperature calculated over a 5×5 pixel window. If the brightness temperature variation does not exceed this threshold, the pixels are considered to belong to the same cloud cluster and are assigned to the cloud growth region.

[0075] The isolated convective cloud sample acquisition unit is used to obtain isolated convective cloud samples based on the cloud growth area.

[0076] The isolated mature convective cloud cluster sample screening module is used to use eight-connected region analysis to distinguish mature convection from immature convection on the isolated convective cloud cluster samples, and screen out the isolated mature convective cloud cluster samples.

[0077] Specifically, after the connected area is determined, short-wave near-infrared and long-wave infrared are used to distinguish between mature convection and immature convection. The near-infrared channel (3.7μm) can reflect the size of cloud top particles. The lower the albedo of this channel, the larger the cloud droplets are, and it is more likely to develop into mature convection. Combined with the brightness temperature of the infrared channel (10.8μm), which reflects the cloud top height, higher clouds are also more likely to produce mature convection. Figure 2 As shown in the figure, the maturity of isolated convective cloud clusters identified by satellites is judged. The judgment criteria include: the cloud top brightness temperature (BTcloud(T)) is continuously below 0℃, and the temperature change before and after is greater than 8℃ (i.e., BTcloud(t)-BTcloud(t-1) ≤ -8℃ and BTcloud(t+1)-BTcloud(t) ≤ -8℃), and the reflectivity of the 3.7 μm channel (Ref3.7) must be greater than 0.40. Only clouds that meet the above three indicators are considered mature convective clouds. Otherwise, the grid pixels are removed.

[0078] The sample dynamic verification module is used to use the pre-processed radar data to verify the convective area of ​​the isolated mature convective cloud cluster sample to obtain the final isolated mature convective cloud cluster sample.

[0079] The sample dynamic inspection module includes:

[0080] The radar echo corresponding cloud mask pixel acquisition unit is used to use the isolated mature convective cloud sample cloud mask extracted from the Fengyun geostationary meteorological satellite data, combine the binary function of the isolated mature convective cloud sample cloud mask and the radar echo area corresponding to the isolated mature convective cloud sample cloud mask, and obtain the number of pixels of the isolated mature convective cloud sample cloud mask corresponding to the radar echo area.

[0081] The cloud mask total pixel acquisition unit is used to obtain the total number of pixels of the isolated mature convective cloud sample cloud mask by using the binary function of the isolated mature convective cloud sample cloud mask.

[0082] The matching rate calculation unit is used to calculate the matching rate between radar echo areas of different intensities and the isolated mature convective cloud sample mask by using the number of pixels of the isolated mature convective cloud sample mask corresponding to the radar echo area and the total number of pixels of the isolated mature convective cloud sample mask.

[0083] Specifically, the extracted samples are dynamically tested in combination with radar. Considering that satellites may observe convection information earlier than radar, regions with different radar echo intensities (15dBZ, 20dBZ, 25dBZ, and 30dBZ) are used to match the cloud mask. The matching rate between the satellite cloud mask and the radar echo is defined as:

[0084] , (4)

[0085] Among them, M d Indicates the satellite observation time t s and radar observation time t r The matching rate between the cloud mask and the radar echo area under the condition of s ) is the satellite at t s The cloud mask binary function observed at the moment is 1 if there is cloud at the pixel (x, y), otherwise it is 0; R d (x, y, t r ) is the radar at t r The observed echo mask at that moment (1 if the intensity exceeds the threshold, 0 otherwise). The numerator represents the number of pixels corresponding to radar echoes that meet the conditions within the cloud mask area. The denominator is the total number of pixels in the cloud mask area.

[0086] The dynamic verification unit is configured to use the isolated mature convective cloud cluster sample whose matching rate meets a preset threshold as the final isolated mature convective cloud cluster sample. In this embodiment, the preset threshold is the matching rate M d >0.5.

[0087] Because factors initiating and triggering convection can both influence its development and evolution, they overlap with the physical factors that characterize its evolution. The development and extinction of convection stem from, firstly, changes in atmospheric dynamic and thermodynamic conditions caused by evolving weather patterns, and secondly, interactions and feedback relationships between convective systems, leading to the rapid growth or extinction of convective populations. While the mechanisms are complex and difficult to describe simply with individual factors, some physical parameters can still predict convection development after it initiates. For example, the water vapor channel with a central wavelength of 6.7 μm is highly sensitive to water vapor radiation in the middle and upper troposphere. By comparing the brightness temperature difference between this channel and the IR window channel, the position and development rate of cumulus cloud tops relative to the middle and upper troposphere can be estimated. Furthermore, visible and near-infrared (0.6 μm-3.9 μm) data are also used to quantitatively describe or invert the evolution of cloud physical characteristics (such as cloud optical thickness, cloud top phase, cloud particle effective radius, and cloud top morphology) during the development of convectively associated cumulus clouds during daytime.

[0088] The recognition model building module is used to fuse and verify the satellite characteristics and radar characteristics of the final isolated mature convective cloud cluster samples and establish an intelligent recognition model.

[0089] The recognition model construction module includes: a sample construction unit, a feature extraction unit and a model construction unit.

[0090] The sample construction unit is used to verify the identification of nascent convection on final isolated mature convective cloud samples using radar data. Final isolated mature convective cloud samples that pass verification are labeled as positive samples, while isolated convective cloud samples that fail to develop into mature convection are labeled as negative samples. By tracking and analyzing the cloud life cycle, information such as its duration and evolution trajectory is extracted, and radar echo intensity is paired with cloud clusters to construct positive and negative sample sets for training the intelligent recognition model for isolated mature convective clouds. This embodiment also includes balancing positive and negative samples based on the number of positive and negative samples, using random sampling strategies, etc., to ensure that the positive and negative sample ratio is no less than 1:2, to prevent the model from being biased towards the mainstream category during training.

[0091] A further embodiment is that the process of using radar data to verify the inception of convection on the final isolated mature convective cloud cluster sample includes:

[0092] The Fengyun geostationary meteorological satellite data and the radar data are time-matched according to the satellite-radar time matching span; the satellite-radar time matching span includes 15 minutes, 30 minutes and 60 minutes; Figure 2As shown in the figure, in radar data, if the reflectivity factor (dBZ) meets any of the following conditions: dBZ ≥ 25 for 15 minutes, dBZ ≥ 35 for 30 minutes, or dBZ ≥ 45 for 60 minutes, a potential convective system is considered to exist; at the same time, ground precipitation observations are combined to enhance the spatiotemporal continuity of identification.

[0093] Specifically, satellites have an advantage in identifying the incipient convection early on. However, to verify whether convection has successfully developed, this technology further incorporates ground-based weather radar observations for verification. The principle is that during the developing and mature stages of convection, radar can observe a large reflectivity factor Z, a strong differential reflectivity factor ZDR, a specific differential phase shift ratio KDP, and a low correlation coefficient (CC). The correlation coefficient is an inversion parameter used in dual-polarization radar. These observations are used to verify the occurrence of convection. For different satellite-radar time matching spans (15 minutes, 30 minutes, and 60 minutes), threshold combination criteria are set within different time windows based on the general physical process of convection development.

[0094] When the satellite-radar time matching span is within 15 minutes, the convection is in the development stage. The observations and the development stage threshold are used to infer the initial identification results of convection in the first 15 minutes. The observations include reflectivity factor, differential reflectivity, correlation coefficient, and differential phase shift ratio.

[0095] When the satellite-radar time match span is within 15-30 minutes, the convection is in the mature stage. The observation value and the mature stage threshold are used to infer the initial convection identification results in the first 30 minutes. The specific initial threshold settings are shown in Table 2 and can be adjusted according to the region later:

[0096] Table 2

[0097]

[0098] When the satellite-radar time match span is within 60 minutes, radar echo characteristics are used to verify the results of convection inception identification. This verification, based on radar echo characteristics within the subsequent time window of identification inception, eliminates false identifications caused by satellite inversion errors and suppressed convection development, significantly reducing false alarm rates.

[0099] Finally, based on satellite-ground verification, a relatively objective final isolated mature convective cloud cluster sample is established as the standard for subsequent intelligent recognition model training and evaluation.

[0100] The recognition model building module is used to integrate and verify the satellite characteristics and radar characteristics of isolated mature convective cloud samples and establish an intelligent recognition model.

[0101] In a further embodiment, the recognition model building module includes:

[0102] The feature extraction unit is used to extract feature parameters of Fengyun geostationary meteorological satellite data and radar data.

[0103] Specifically, input features include not only traditional information such as satellite channel brightness temperature, brightness temperature difference, cooling rate, and cloud microphysical characteristics (such as liquid water path (LWP) and ice water content (IWC)), but also radar characteristic parameters and environmental parameters (such as CAPE and CIN) from multiple sources. Radar characteristic parameters include the maximum reflectivity factor derived from the reflectivity factor. The extraction process involves finding the maximum reflectivity factor at each altitude of the convection location.

[0104] The model building unit is used to train a gradient boosting decision tree using the characteristic parameters to construct an intelligent recognition model. Specifically, the binary classification model XGBClassifier is selected, the loss function is the error rate 'error', and the hyperparameters are set: 10-100 weak learners, a maximum depth of 6-10, a learning rate of 0.01-0.1, and a sample ratio of 0.8-1.0 per learning session. The accuracy metric is set to evaluate model performance, and the model is saved after training. This model combines multiple characteristic information such as cloud top brightness temperature, brightness temperature change rate, reflectivity, and cloud geometry to achieve efficient identification of isolated mature convective clouds, improving the early identification of potential severe convective events.

[0105] In this embodiment, a threshold method is provided to iteratively improve the model:

[0106] The output results of the intelligent recognition model are compared and analyzed with the original empirical threshold-based method.

[0107] If the model performs better than traditional methods, it can be used to reversely guide the optimization and adjustment of the empirical threshold based on the analysis of the importance of model features, and gradually form a more intelligent and dynamic recognition system.

[0108] The training set is continuously updated by regularly introducing new samples and online learning during user use, so as to achieve continuous iterative upgrades of the model.

[0109] The recognition module is used to dynamically identify isolated mature convective cloud clusters based on an intelligent recognition model. Ultimately, this system forms a detection system for isolated mature convective cells that combines physical law judgment with data-driven intelligent recognition.

[0110] This implementation integrates FY4A / AGRI and surface minute-by-minute precipitation data to automatically track and identify convective systems. By comparing the combined reflectivity with Doppler weather radar data, the algorithm can provide a 1.5-hour lead time for convection onset signals. Based on FY4A / AGRI near-infrared and water vapor channels, the algorithm automatically extracts isolated mature convective clouds, which align with the spatial locations of strong echoes detected by Doppler radar.

[0111] Example 2:

[0112] The present invention also provides a method and application system for dynamically identifying isolated mature convective clouds by integrating satellite and radar, including:

[0113] Preprocess the acquired Fengyun geostationary meteorological satellite data and radar data.

[0114] The preprocessed Fengyun geostationary meteorological satellite data are used to obtain the cloud mask of convective clouds, and the cloud mask is used to identify convective cloud cluster samples.

[0115] Based on the convective cloud samples and combined with the cloud growth area, isolated convective cloud samples are determined.

[0116] Eight-connected region analysis is used to distinguish mature convection from immature convection in isolated convective cloud clusters, and isolated mature convective cloud clusters are selected.

[0117] The preprocessed radar data are used to verify the convective area of ​​the isolated mature convective cloud cluster samples and obtain the final isolated mature convective cloud cluster samples.

[0118] The satellite and radar characteristics of the final isolated mature convective cloud samples are fused and verified to establish an intelligent recognition model.

[0119] Based on the intelligent recognition model, dynamic identification of isolated mature convective cloud clusters is completed.

[0120] In a further embodiment, the method for identifying isolated convective cloud samples includes:

[0121] Cold cloud cores are identified using the infrared channel of preprocessed Fengyun geostationary meteorological satellite data.

[0122] The growth search method is used to check the adjacent pixels of the corresponding pixel of the cold cloud core to determine whether the adjacent pixels meet the cloud growth conditions and obtain the cloud growth area.

[0123] Based on the cloud growth area, isolated convective cloud samples are obtained.

[0124] In a further embodiment, the method for obtaining a final isolated mature convective cloud sample comprises:

[0125] The sample cloud mask of isolated mature convective cloud clusters extracted from Fengyun geostationary meteorological satellite data was used. The number of pixels of the sample cloud mask of isolated mature convective cloud clusters corresponding to the radar echo area was obtained by combining the binary function of the sample cloud mask of isolated mature convective cloud clusters and the radar echo area corresponding to the sample cloud mask of isolated mature convective cloud clusters.

[0126] The total number of pixels of the sample cloud mask of isolated mature convective cloud clusters is obtained using the binary function of the sample cloud mask of isolated mature convective cloud clusters.

[0127] The matching rates between radar echo areas of different intensities and the sample cloud masks of isolated mature convective clouds are calculated using the number of pixels of the sample cloud masks of isolated mature convective clouds corresponding to the radar echo areas and the total number of pixels of the sample cloud masks of isolated mature convective clouds.

[0128] The isolated mature convective cloud samples whose matching rate meets the preset threshold are taken as the final isolated mature convective cloud samples.

[0129] The embodiments described above are merely descriptions of preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by persons skilled in the art should fall within the scope of protection defined by the claims of the present invention.

Claims

1. A dynamic identification system for isolated mature convective clouds integrating satellite and radar, characterized by: include: Data processing module, used to pre-process the acquired Fengyun geostationary meteorological satellite data and radar data; a convective cloud cluster sample extraction module, configured to obtain a cloud mask of convective clouds using pre-processed Fengyun geostationary meteorological satellite data, and identify convective cloud cluster samples using the cloud mask; an isolated convective cloud sample identification module, configured to determine the isolated convective cloud sample based on the convective cloud sample and in combination with the cloud growth area; An isolated mature convective cloud cluster sample screening module is used to use eight-connected region analysis to distinguish mature convection from immature convection on the isolated convective cloud cluster samples, and screen out isolated mature convective cloud cluster samples; a sample dynamic verification module, configured to verify the convective area of ​​the isolated mature convective cloud cluster sample using the preprocessed radar data to obtain a final isolated mature convective cloud cluster sample; The recognition model building module is used to integrate and verify the satellite characteristics and radar characteristics of the final isolated mature convective cloud cluster samples and establish an intelligent recognition model; The recognition module is used to complete the dynamic recognition of isolated mature convective cloud clusters based on the intelligent recognition model.

2. The system according to claim 1, wherein: The isolated convective cloud sample identification module includes: The cold cloud core identification unit is used to identify cold cloud cores using the infrared channel of the pre-processed Fengyun geostationary meteorological satellite data; a cloud growth region acquisition unit, configured to use a growth search method to check adjacent pixels of a pixel corresponding to the cold cloud core, determine whether the adjacent pixels meet cloud growth conditions, and obtain a cloud growth region; The isolated convective cloud sample acquisition unit is used to obtain isolated convective cloud samples based on the cloud growth area.

3. The system according to claim 2, characterized in that The sample dynamic inspection module includes: The radar echo corresponding cloud mask pixel acquisition unit is used to use the isolated mature convective cloud sample cloud mask extracted from the Fengyun geostationary meteorological satellite data, combine the binary function of the isolated mature convective cloud sample cloud mask and the radar echo area corresponding to the isolated mature convective cloud sample cloud mask, and obtain the number of pixels of the isolated mature convective cloud sample cloud mask corresponding to the radar echo area; a cloud mask total pixel acquisition unit, configured to obtain the total number of pixels of the isolated mature convective cloud sample cloud mask by using a binary function of the isolated mature convective cloud sample cloud mask; a matching rate calculation unit, configured to calculate matching rates between radar echo regions of different intensities and the isolated mature convective cloud sample mask using the number of pixels of the isolated mature convective cloud sample mask corresponding to the radar echo region and the total number of pixels of the isolated mature convective cloud sample mask; The dynamic verification unit is configured to take the isolated mature convective cloud cluster sample whose matching rate meets a preset threshold as the final isolated mature convective cloud cluster sample.

4. The system according to claim 1, wherein: The recognition model building module includes: A sample construction unit is used to verify the identification of the incipient convection of the final isolated mature convective cloud cluster samples using radar data, marking the final isolated mature convective cloud cluster samples that pass the verification as positive samples, and marking the isolated convective cloud cluster samples that fail to develop into mature convection as negative samples; A feature extraction unit is used to extract feature parameters of Fengyun geostationary meteorological satellite data and radar data; The model building unit is used to use the characteristic parameters to train the gradient boosting decision tree and build an intelligent recognition model.

5. The system according to claim 4, characterized in that The process of using radar data to verify the identification of convective inception of the final isolated mature convective cloud cluster sample includes: Time-matching of Fengyun geostationary meteorological satellite data with radar data using satellite-radar time matching spans of 15 minutes, 30 minutes, and 60 minutes. When the satellite-radar time matching span is within 15 minutes, the convection is in the development stage, and the observations and the development stage threshold are used to infer the initial convection identification results of the previous 15 minutes; wherein the observations include reflectivity factor, differential reflectivity, correlation coefficient, and differential phase shift ratio; When the satellite-radar time matching span is within 15-30 minutes, the convection is in the mature stage, and the observations and the mature stage threshold are used to infer the initial convection identification results of the first 30 minutes; When the satellite-radar time matching span is within 60 minutes, the radar echo characteristics are used to verify the convection initiation identification results.

6. A method for dynamically identifying isolated mature convective clouds by integrating satellite and radar, using the system according to any one of claims 1 to 5, characterized in that: include: Pre-process the acquired Fengyun geostationary meteorological satellite data and radar data; Using pre-processed Fengyun geostationary meteorological satellite data, a cloud mask of convective clouds is obtained, and the convective cloud cluster samples are identified using the cloud mask; Based on the convective cloud samples, and in combination with the cloud growth area, an isolated convective cloud sample is determined; Eight-connected region analysis is used to distinguish mature convection from immature convection on the isolated convective cloud cluster samples, and isolated mature convective cloud cluster samples are screened out; Using the pre-processed radar data to verify the convective area of ​​the isolated mature convective cloud cluster sample, and obtaining a final isolated mature convective cloud cluster sample; The satellite and radar characteristics of the final isolated mature convective cloud cluster samples were integrated and verified to establish an intelligent recognition model; Based on the intelligent recognition model, dynamic recognition of isolated mature convective cloud clusters is completed.

7. The method according to claim 6, characterized in that The method for identifying the isolated convective cloud sample includes: Cold cloud cores are identified using the infrared channel of pre-processed Fengyun geostationary meteorological satellite data. Adopting a growth search method to check the adjacent pixels of the pixel corresponding to the cold cloud core, determining whether the adjacent pixels meet the cloud growth conditions, and obtaining the cloud growth area; Based on the cloud growth area, isolated convective cloud samples are obtained.

8. The method according to claim 7, characterized in that The methods for obtaining the final isolated mature convective cloud sample include: The isolated mature convective cloud sample mask extracted from Fengyun geostationary meteorological satellite data is used. The number of pixels of the isolated mature convective cloud sample mask corresponding to the radar echo area is obtained by combining the binary function of the isolated mature convective cloud sample mask and the radar echo area corresponding to the isolated mature convective cloud sample mask. Using a binary function of the isolated mature convective cloud sample cloud mask, a total number of pixels of the isolated mature convective cloud sample cloud mask is obtained; The matching rates between radar echo areas of different intensities and the isolated mature convective cloud sample masks are calculated using the number of pixels of the isolated mature convective cloud sample mask corresponding to the radar echo area and the total number of pixels of the isolated mature convective cloud sample mask. The isolated mature convective cloud cluster sample whose matching rate meets a preset threshold is used as the final isolated mature convective cloud cluster sample.

Images